/diy/ - Do It Yourself

car battery charger

take out all the fans and make a huge wind tunnel

>>I was thinking turning one into a workbench psu; any tips for that?
Just short the green wire to any black wire on the ATX connector.

Yeah, make a bench top power supply with one for sure (I have one from ages ago - works a charm)

For the rest... strip them down, pull out the capacitors and make a coil gun.

ATX PSU's have some cool bits in, worth having a few stripped and sorted for their transformers, caps etc. Although watch out for bulging/leaking caps - just throw them out.

http://www.wikihow.com/Convert-a-Computer-ATX-Power-Supply-to-a-Lab-Power-Supply

but with more tweaks, you can make something like that

fan?

> I was thinking turning one into a workbench psu; any tips for that?

Yes, DONT FUCKING DO IT.

everybody makes mistakes as they're experimenting. if you make a mistake and your power supply can source 20 amperes, minor mistakes become major mistakes. your breadboard will melt and get charred. your alligator connections will melt and burn. transistors and capacitors will explode. your tools will end up with markings will metal has evaporated from a short circuit. your fingers will get burned.

Never use a 20A power supply while experimenting. you'll become gun shy, so you wont have any fun, and you'll waste a lot of money replacing burnt out parts.

A current-limiting power supply can be your best friend.

>>212384
Other than that, the PSU used in PCs are not very good material for lab/bench supplies other than being cheap and available.

The overcurrent protection kicking in about 50A range is one thing. The others are poor voltage stability in most cases (especially for cheap models) and serious ripple that needs filtering for any analog circuitry. The PSU might not even start without some minimal load or behave strangely. It might "not like" the load being only on 5V or 12V bus, also just one voltage is probed for regulation - usually the lowest positive that the PSU has. So 3V3 in this case. The rest is stabilized only to match the probed voltage. In other words, a serious load on 12V line can drop it to less than 11V if the 3V3 is completly unused. It's only thanks to all windings being on one transformer that it doesn't get any worse.

And, lastly, a PC PSU must be grounded properly. Failure to observe this rule will result in electric shocks - that don't kill per se, but easily damage electronics. It's because of how the input capacitors are wired.

>>212368
>make a coil gun.

A really really tiny one.

>>212384
>>212417
Wow, you guys know only enough about PSUs to look stupid. lol

Use the 5v and 12v rails to power things. You can wire up small car subwoofers to them if the 12v rail pushes enough amps and hook that up to your computer or home theater system.

>your breadboard will melt and get charred. your alligator connections will melt and burn. transistors and capacitors will explode. your tools will end up with markings will metal has evaporated from a short circuit. your fingers will get burned.

>>212384
>>212417

Seriously, is that guy telling the truth?

>>212565
It is true that a PSU requires a load for proper regulation, it is also true that they can deliver a lot of current.

IDGAF though, I still use them in a pinch.

>>212565
If you have a multimeter and a clamp meter you should do fine. The volts and amps are written on the side of the PSU. It's not rocket science. Use a transformer and/or resistors to get it to what you need.

>>212565
EE student here

they can supply a lot of current, its mostly on the 3v and 5v rails.

what i did was use cheap automotive fuses on all the outputs. 1amp fuses are more then enough for anything electronics wise.

send them to us, I'm in need of an ATX power supply.

>Collect more PSUs
>make huge pile
>????
>PROFIT!!!

OP picture kinda remind me of that.

>>212354
> take out fans
> make wind tunnel

this is the best suggestion in this entire thread

>>212435
How does one do this?

>>212579
Also EE student..

> 1amp fuses are more then enough for anything electronics wise.

Study harder my young padawan, and you'll see a world of electronics that requires a lot more than 1A to even start.

>>212626
What game is that?

>>213098
I was going to answer that post too, but then I realized doing so would be retarded.

>>213100


Lookes like dues ex revolution

Muck about with them to turn them into "home anodizing kits" sell on ebay.

>>212349
The only problem with using PC supplies as a bench supply is many of them require a minimum load on one of the outputs (usually the +5 or the +3.3V) to function. That, and the lack of adjustability.

>>213698
Just slap a 10W 10Ohm resistor between the 5V line and ground. That's what I did. Hell the website OPs pic came from says to do that:
http://web2.murraystate.edu/andy.batts/ps/powersupply.htm

Google: how to modify old power supplies to convert them to 650watt for use with highend video cards, and sell them to children on ebay.

step one print fancy labels that say 650W*
step two attach label
step three profit

>>212349
as a scrap commodity, they go for about 20cents per pound, in bulk as is. they contain steel, copper, aluminum, etc. mixed. If you take a hammer to them, transformers, and other copper coils sell for more.
if you have one of these...

http://www.youtube.com/watch?v=sQYPCPB1g3o

and a magnetic separator, you can increase your profitability

>>212579

This.

"Hurr durr" no current regulation faggots dont seem to understand that you can do it quite happily yourself!

http://en.wikipedia.org/wiki/Current_source#Simple_transistor_current_sources

You'll need a mosfet and some high power resistors, though for most electronics applications I'd consider 1Amp sufficent. Still, that's probably a 30 Watt tolerance for +/12 Volt lines. You can then either use a Voltage Controlled resistor (another MOSFET with a DC voltage driving the Drain) and a trimpot or a high power selector switch with various resistor values to vary the output current limit.

>>213733 stupid fag doesnt know there's a huge difference between current regulation and current limiting. also doesnt understand it would require one current limiting circuit per voltage. or that it would affect the output voltages, so 5V and 12V which are desirable voltages for experimenting with would become undesirable 3.5V and 10.5V voltages, for example.

take em to the scrap yard..you can get good money for them

>>213862

Learn to greentext.

Then learn to brain. If it's a constant input voltage, then the differences are arbitrary, dickhead. Even if you should be putting both terminals to your eyeballs whilst theirs a sag on the input voltage a series resistor will limit how far this madness continues and, if not, the fuse should blow.

Thirdly, you just take your 10.5 and regulate it down to 9 and 5. O.k, so you don't have +/- 12 volts of DC, your circuit should still work fine with +/- 9.

Unless your a dickhead and can't design for shit.

>>213703

I have two, I use them to run my battery chargers for my R/C planes and cars.

>>213733
Perfect example of knowing very little and being confident knowing very little.

Converted ATX power supplies make bad bench supplies.

Things you want in a bench supply:

1) +/- 0-15 v dual tracking rails (500ma+)

2) Fast acting current limiting.

3) Clean DC voltage

The thing about switching supplies is that generally relatively high frequency ripple will make it through to the output and need attenuation. High frequency noise is more annoying to deal with because many opamps have worse PSRR at higher frequencies.

With a good quality power supply, the ripple will probably be tolerable, but those power supplies look like garbage.

Regulation is also questionable (as people have said).

240/110v extension cord powered hovercraft

it sends them to me or it gets the hose again

It takes very little effort to remove the steel casing from a PSU, as it's typically only screws holding components to it; maybe the occasional zip-tie inside.

You can clip the wires easily to take those in for recycling as well.

If you want to go a little further, it takes a touch more effort (sometimes you have to pull things off the board to reach the screws in this case), you can unscrew components from the aluminum heatsinks and pull those off as well.

Take the rest to an E-Waste drop off or see if someone will take them for cash as well; depends on what's available in your area.

>>213098

Then he takes out the 1A and replaces it with 5 or 10 or whatever he needs...?

/diwhyamIstillhere/

>>212435

If the OP has anywhere near the number of PSU's pictured he could make a pretty sizeable coilgun capacitor bank from the power caps.

Unless he's in the US? I guess they will be ~200 V there. In the UK they are often ~400 V

>>214169
The biggest problem with getting stuff off the PCB is this glue gunk they put on everything in some PSUs. Sometimes it'll pop right off and other times it's there for good and you have to cut or even grind it off. It can be pretty terrible.

I have a broken PSU... it'll turn on but then immediately starts burning and smoking and the fan spins like mad.

Worked fine for five years then just died.

What can I MAKE (not recycling for cash) out of this? I am not OP

>>214201
If you have a multitester you can find the problem and fix it, or at least know when component to not save and take the rest for EE projects.

> If you have a multitester you can find the problem and fix it,

nope, even experienced people have trouble with switching power supplies because the way they work is that everything is inter-related. so isolating a non-obvious problem is hit-and-miss.

>>214201
pull out all the torroid cores and make a bunch of joule thiefs.

http://www.youtube.com/watch?v=gTAqGKt64WM

>>214199

Heat gun. No, really, they're great for that stuff. If you've ever seen someone replace the transformer in an SM57 (which is a giant tube of fuckin' epoxy) then you'll see what I mean. Good for surface mount stuff, too. It's on my to-buy-list!

>>214201

Well if you've done that a few times then chances are everythings fucked. May have been bad/leaky caps or a leaky rectifier, may have been that the onboard PIC/IC suffered a degredation from the heat caused by fast switching speeds. Initial start up is the highest risk moment for this - you really need to just leave the higher current ones ON.

>>214245
>Heat gun.

That's good only for the stuff that isn't dried out. I've tried it before.

power a railgun?